It’s Time for a Conversation About A.I. Use in Flint

Header image by Justin Doverspike

I promise this isn’t about a fox. Well, it isn’t entirely about a fox.

Since launching Flint. Daily., I’ve periodically snuck in my thoughts about the growing reliance on faulty artificial intelligence technology being deployed to create worse versions of things that humans have been doing a perfectly fine job of making (hint: I think it sucks!). A couple months ago, I wrote a column to articulate some of my higher level frustrations with A.I. To recap:

• Regulation of the industry and how the technology is used is light years behind how fast it is being relied on, resulting in horrendous problems like mass proliferation of child sexual abuse material (CSAM) and deepfakes. If that sentence alone isn’t enough to disgust you, 404 Media just published a thoroughly reported story about how five teenage girls at a Pennsylvania high school had their lives upended by how easy this material is to produce, and how ill-equipped law enforcement is to detect and prevent it.
• The environmental impact is at best unclear and at worst potentially disastrous as data centers are rapidly constructed to support the technology, and vital things like our power grid and water supplies are potentially impacted in harmful ways.
• The economic prospects are dubious. When companies are throwing around fake sounding economic projections (like billionaire team owners that want taxpayer funding for stadiums that never provide the promised return on investment), that’s a major red flag. There’s not solid evidence showing that data centers and A.I. will produce the promised economic windfalls, tax revenue, or mass employment after they are constructed.

Those are just high-level, existential things though. This is about how A.I. has a negative impact locally, and why the many businesses, politicians, students, and other entities using it should reconsider. And I’m going to start with two areas that I work in as examples.

This very website has a simple A.I. statement on our About page. The gist: we don’t use it. For anything.

MLive, on the other hand, has made the decision as a company to use it. Anything with the byline ‘Advance Local Express Desk‘ uses A.I. to create content you are reading in that article. Up front I want to stress this isn’t about individual journalists who work at MLive. I know several and have good relationships with them and respect their work. Their company making what I believe is an unethical decision to introduce A.I. into the newsgathering process is a mistake, though.

That decision, like many A.I. decisions, was likely made with workload in mind — they can feed press releases or information to their A.I. tool and churn out simple stories that humans in their shrunken newsrooms may not have otherwise been able to get to. That concept defeats the purpose of what we do, though. News is as much about relationships and talking to people as it is about whatever outputs come out of that. Flint. Daily. is not my full-time job, and I could certainly produce more content by using A.I. I understand the appeal of wanting to tell more stories in a shorter amount of time.

But that defeats the core purpose of what I’m trying to do here. I care more about spending time talking with people in the community than whether or not those conversations always lead to a story I can crank out and post. I want people who I write about to know that I do so with care, with intention, and with a real love for this city and living here in it. I can’t convey that sentiment by using a tool that strips all love for the process out of community journalism in an effort to publish an extra press release or two every day.

Consumers build relationships with writers, artists, photographers, and other creative people because they bring distinctive voices and styles to their work. I don’t know if I succeed with that at Flint. Daily., but sounding like ‘me’ is definitely my goal. If publications all start using A.I. in the interest of volume, and everything sounds the same, what reason do people have to seek out my work if it just becomes part of a milquetoast sea of boringness?

I also teach. And if you’re among the A.I. evangelists out there, I’m begging you to spend time talking with any teacher in a classroom about the negative ways A.I. impacts learning. I’ve taught communications and public speaking both on a college campus and in a prison. In a prison setting, you obviously have less access to technology, so I’ve always used more of an analog approach there — pencils and paper, writing in class rather than at home, etc. The last time I taught on campus, after rampant A.I. use resulted in a bunch of students giving the same boring speeches, I pivoted and brought the “let’s do everything right here together” approach to that class and stopped using most technology in class and moved writing from homework to in class. For two weeks, the students hated it. By the end of the semester, they loved it — it gave them a chance to re-exercise all of the thinking muscles that A.I. usage is killing in us.

I am fortunate enough to teach courses that can easily ditch most technology. For teachers who have a harder time adding safeguards, combatting students constantly cheating on assignments — and that’s exactly what using A.I. is — has become a second full-time job. And for the students who are able to skate through or not have their cheating detected, what’s the reward? Getting a degree you didn’t earn? Finishing a class where you didn’t absorb or learn any of the materials? Getting into a career field where you’re not actually prepared? Who is any of that helping? At least the isolated weirdos giving commencement addresses about the wonders of A.I. keep getting booed off the stage by students who hate it.

The Ethics of Selling Fake Hamburgers

We’ll get to ugly flyers that all look the same in a second. Before that, I want to call out one component of those flyers. If you own a restaurant or other food-serving establishment and you use A.I. to create your marketing materials, and those materials include A.I. pictures of food that you don’t cook or sell, explain how that’s ethical marketing?

“Here’s some cartoon food that doesn’t look like anything we serve!”

Technically, advertising a product you don’t actually sell is against Federal Trade Commission (FTC) regulations. I don’t really understand how A.I.-generated images of products that aren’t real don’t violate that provision.

But beyond that, the images people use often look worse than the products they’re really selling! Just take pictures of your real food. The uniqueness is the appeal for any small business, and using A.I. is stripping you of your greatest asset.

Get A.I. Off Your Campaign Materials

Consider this a callout not aimed at any specific local candidate and rather a blanket statement: if I see campaign literature from you as a candidate that looks A.I. generated, it’s disqualifying for me as a voter. I can’t support a candidate who uses it.

I’ve seen some flyers posted on social media that have fake people in them, fake imagery that is supposed to be places in Flint, and clearly uses the famous overpowering font that most people now immediately associate with an A.I. flyer. Using A.I. to include photos of fake people or distorted representations of Flint iconography on a flyer telegraphs what you’ll be like as a representative: lazy. Walk around in Flint and take photos with your supporters or at events you attend. Walk around the city and take pictures of the landmarks. Use those on your flyers.

Incidentally, Michigan law requires that candidates disclose if A.I. is used to create all or a portion of a piece of campaign literature or materials. And … let’s just say I have not noticed many of those disclosures on things that are clearly made with A.I.

A.I. Defeats the Purpose of Branding

The point of marketing and branding is to differentiate your business or yourself as an entity. Using a tool that makes everything look the same, and one that uses bad design principles at that, has the opposite effect.

I have sympathy for some people and organizations that use it. Nonprofit organizations that are strapped for resources and time, or even governmental bodies trying to save money are reasonable excuses — to bring this back to the fox conversation, I understand why the city of Flint doesn’t want to cause controversy by spending what an animator might charge for a Flinny the Fox video. I still think it’s the wrong approach — as reader and local artist Kristina Lakey pointed out in the comments of our Facebook post, if done transparently, paying a local artist to come up with a design and concept for a character could be perceived positively.

I think it’s the wrong approach for small businesses and nonprofits, too. Even if they don’t have budget for a designer, or the skillset in-house, the uniformity of A.I. flyers and the increasingly negative reaction to A.I.-generated materials are bad for branding.

I also empathize with people who, without budget, are intimidated to try and design their own work, so they turn to A.I. But I assure you, the quaint or amateurish flyers that were common pre-A.I. are far superior to the overpowering, graphic-heavy, sometimes disfigured soulless creations that A.I. comes up with. Doing something quaint and amateurish brings a charm with it that is endearing, and ensures whichever customer base you’re trying to reach that there’s a human behind the work you do.

Even putting aside the massive ethical concerns, A.I. is error-prone, has an increasingly negative reputation (especially among young people), and from a design perspective, it doesn’t offer anything better or more professional or unique than a novice using the free version of Canva could come up with on their own in roughly the same amount of time. The faster we rethink how much we are using it for tasks we can do ourselves, the better off we will be.